README.md

metadata

configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
dataset_info:
  features:
    - name: sample_id
      dtype: string
    - name: layers
      list:
        image:
          decode: false
    - name: preview
      dtype:
        image:
          decode: false
    - name: rendered
      dtype:
        image:
          decode: false
    - name: boundingbox
      struct:
        - name: format
          dtype: string
        - name: boxes
          list:
            list: float32
        - name: meta
          dtype: string
  splits:
    - name: train
      num_bytes: 32521424020
      num_examples: 19479
  download_size: 31847217475
  dataset_size: 32521424020

---

task_categories: - image-segmentation task_ids: - semantic-segmentation pretty_name: DLCV Final Dataset size_categories: - medium

DLCV Final Dataset

This dataset is used for the Deep Learning for Computer Vision (DLCV) final project.
It contains ground-truth layers organized per sample and is designed for training and evaluating computer vision models.

---

📂 Dataset Structure

The dataset is organized as follows:

dlcv_final/

├── gt_layers/

│ ├── sample_0000/

│ │ ├── layer_0.png

│ │ ├── layer_1.png

│ │ └── ...

│ ├── sample_0001/

│ ├── sample_0002/

│ └── ...

└── README.md

• Each sample_xxxx directory corresponds to one data sample
• Files inside each sample directory represent ground-truth layers
• Folder structure is preserved to simplify indexing and loading

---

🚀 How to Use

You can access this dataset using the 🤗 datasets library:

from datasets import load_dataset

dataset = load_dataset("dereklin1205/dlcv_final_dataset")